seqcounter 1.0.0__tar.gz

seqcounter-1.0.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2022 Marc-Olivier Duceppe
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

seqcounter-1.0.0/MANIFEST.in

@@ -0,0 +1 @@
+recursive-include tests *.py

seqcounter-1.0.0/PKG-INFO

@@ -0,0 +1,140 @@
+Metadata-Version: 2.4
+Name: seqcounter
+Version: 1.0.0
+Summary: Count the occurrences of reference sequences in sequencing reads or genome assemblies
+Author-email: Marc-Olivier Duceppe <marc-olivier.duceppe@inspection.gc.ca>
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/duceppemo/seqcounter
+Project-URL: Documentation, https://github.com/duceppemo/seqcounter/wiki
+Project-URL: Changelog, https://github.com/duceppemo/seqcounter/wiki/Changelog
+Project-URL: Issues, https://github.com/duceppemo/seqcounter/issues
+Keywords: bioinformatics,kmer,counting,reference sequences,fastq,fasta,wgs,BBTools,Seal
+Classifier: Development Status :: 5 - Production/Stable
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Operating System :: MacOS
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: test
+Requires-Dist: pytest>=7; extra == "test"
+Requires-Dist: pytest-cov>=5; extra == "test"
+Dynamic: license-file
+
+<p align="center">
+  <picture>
+    <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/duceppemo/seqcounter/master/assets/logo_dark.svg">
+    <img src="https://raw.githubusercontent.com/duceppemo/seqcounter/master/assets/logo.svg" width="520" alt="seqcounter: count reference sequences in reads or assemblies">
+  </picture>
+</p>
+
+<p align="center">
+  <a href="https://github.com/duceppemo/seqcounter/actions/workflows/tests.yml"><img src="https://github.com/duceppemo/seqcounter/actions/workflows/tests.yml/badge.svg" alt="Tests"></a>
+  <a href="https://codecov.io/gh/duceppemo/seqcounter"><img src="https://codecov.io/gh/duceppemo/seqcounter/graph/badge.svg" alt="Coverage"></a>
+  <a href="https://github.com/duceppemo/seqcounter/releases/latest"><img src="https://img.shields.io/github/v/release/duceppemo/seqcounter?cacheSeconds=3600" alt="Release"></a>
+  <a href="https://pypi.org/project/seqcounter/"><img src="https://img.shields.io/pypi/v/seqcounter?cacheSeconds=3600" alt="PyPI"></a>
+  <img src="https://img.shields.io/badge/python-3.10%E2%80%933.14-blue" alt="Python 3.10–3.14">
+  <a href="LICENSE"><img src="https://img.shields.io/github/license/duceppemo/seqcounter" alt="License"></a>
+  <a href="https://github.com/duceppemo/seqcounter/wiki"><img src="https://img.shields.io/badge/docs-wiki-informational" alt="Documentation"></a>
+</p>
+
+Count how many sequencing reads (fastq) or contigs (fasta) contain each sequence of a reference file, with
+[Seal](https://sourceforge.net/projects/bbmap/) from BBTools. Point it at a reference fasta file and a sample, or a
+folder of samples, and it writes one table of counts per reference sequence and per sample.
+
+Any set of short sequences works: spoligotype spacers, marker or resistance genes, primer or probe sequences,
+plasmid or phage fragments, barcodes, spike-ins. seqcounter does not interpret them; it counts them, reproducibly
+and in the same way for reads and for assemblies.
+
+```
+$ seqcounter -ref refs.fasta -i samples/ -o results/
+Reference   assembly  no_match  paired  single_end
+absent      0         0         0       0
+present     1         0         86      34
+twice       1         0         154     65
+```
+
+## Features
+* **Reads or assemblies**: single-end or paired-end fastq, or fasta, gzipped or not. Both mates of a pair are
+  counted, once each.
+* **Batch mode**: point it at a folder; fastq and fasta files are found recursively and R1/R2 files paired
+  automatically. Several samples can be counted at the same time.
+* **Both strands, and mismatches**: a reference is found on either strand, and with up to `-m` mismatches per k-mer
+  (one by default).
+* **Three reports**: the counts table (`refseqcounts.tsv`, the table the first version of seqcounter wrote, with the
+  command that produced it on its first line), a per-sample summary of how much was read and how much matched, and a
+  JSON report with the parameters, the software versions and the counts.
+* **Nothing is hidden**: references Seal cannot find (shorter than the k-mer size), names that would collide, and
+  samples that failed are reported, not silently dropped. A failed sample leaves the others readable.
+* **Nothing to configure**: no database, no index, no reference genome. Only the Python standard library, plus
+  BBTools.
+* **Tested**: 105 tests, 100% coverage, end to end on synthetic references, genomes and reads with known counts.
+
+## Installation
+```
+conda create -n seqcounter -c conda-forge -c bioconda bbmap python">=3.10"
+conda activate seqcounter
+pip install seqcounter
+```
+BBTools (which provides `seal.sh`) is not on PyPI, so it comes from bioconda; seqcounter itself needs nothing else.
+See [Installation](https://github.com/duceppemo/seqcounter/wiki/Installation) for other options, including running
+it from a clone with `python seqcounter.py`, as the first version did.
+
+## Quick start
+```
+seqcounter -ref refs.fasta -r1 sample_R1.fastq.gz -r2 sample_R2.fastq.gz -o results/  # Paired-end reads
+seqcounter -ref refs.fasta -r1 sample.fastq.gz -o results/                            # Single-end reads
+seqcounter -ref refs.fasta -r1 assembly.fasta -o results/                             # An assembly
+seqcounter -ref refs.fasta -i samples/ -o results/ -p 4                               # A folder, 4 at a time
+```
+`-ref` is a fasta file of the sequences to count: each one gives a row, named after the first word of its `>` line.
+The counts are printed and saved in `results/`. To see it work on data with known counts, run
+`bash examples/run_example.sh`, which generates its own references and samples.
+
+**What a count is**: the number of reads, or contigs, that contain the reference sequence, not the number of copies
+it has in them. A reference present twice in a single-contig assembly is counted once; in reads, about twice as many
+reads match it. See [How it works](https://github.com/duceppemo/seqcounter/wiki/How-it-works).
+
+## Documentation
+The **[wiki](https://github.com/duceppemo/seqcounter/wiki)** covers
+[installation](https://github.com/duceppemo/seqcounter/wiki/Installation),
+[usage and options](https://github.com/duceppemo/seqcounter/wiki/Usage),
+[output files](https://github.com/duceppemo/seqcounter/wiki/Output-files),
+[how it works](https://github.com/duceppemo/seqcounter/wiki/How-it-works),
+[troubleshooting](https://github.com/duceppemo/seqcounter/wiki/Troubleshooting) and the
+[FAQ](https://github.com/duceppemo/seqcounter/wiki/FAQ).
+
+## Citing
+If you use seqcounter, please cite it and BBTools, which does the counting:
+
+> Duceppe M-O. seqcounter: count the occurrences of reference sequences in sequencing reads or genome assemblies.
+> https://github.com/duceppemo/seqcounter
+
+> Bushnell B. BBTools. https://sourceforge.net/projects/bbmap/
+
+GitHub's **"Cite this repository"** button gives the citation in APA and BibTeX formats, from
+[`CITATION.cff`](CITATION.cff).
+
+## Related
+* [spoligotyper](https://github.com/duceppemo/spoligotyper) — *in silico* spoligotyping of the *M. tuberculosis*
+  complex. It counts the 43 spacers the way seqcounter counts any reference, and then reads them as a spoligotype,
+  with the species, the lineage and a PDF report.
+* [brucetyper](https://github.com/duceppemo/brucetyper) — *in silico* Bruce-ladder, Suis-ladder, AMOS and RB51
+  junction PCR for *Brucella*, from assemblies.
+
+## Contributing
+Bug reports, questions and pull requests are welcome: see [CONTRIBUTING.md](CONTRIBUTING.md).
+
+## Author
+Marc-Olivier Duceppe, Canadian Food Inspection Agency (CFIA): marc-olivier.duceppe@inspection.gc.ca
+
+## License
+[MIT](LICENSE)

seqcounter-1.0.0/README.md

@@ -0,0 +1,109 @@
+<p align="center">
+  <picture>
+    <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/duceppemo/seqcounter/master/assets/logo_dark.svg">
+    <img src="https://raw.githubusercontent.com/duceppemo/seqcounter/master/assets/logo.svg" width="520" alt="seqcounter: count reference sequences in reads or assemblies">
+  </picture>
+</p>
+
+<p align="center">
+  <a href="https://github.com/duceppemo/seqcounter/actions/workflows/tests.yml"><img src="https://github.com/duceppemo/seqcounter/actions/workflows/tests.yml/badge.svg" alt="Tests"></a>
+  <a href="https://codecov.io/gh/duceppemo/seqcounter"><img src="https://codecov.io/gh/duceppemo/seqcounter/graph/badge.svg" alt="Coverage"></a>
+  <a href="https://github.com/duceppemo/seqcounter/releases/latest"><img src="https://img.shields.io/github/v/release/duceppemo/seqcounter?cacheSeconds=3600" alt="Release"></a>
+  <a href="https://pypi.org/project/seqcounter/"><img src="https://img.shields.io/pypi/v/seqcounter?cacheSeconds=3600" alt="PyPI"></a>
+  <img src="https://img.shields.io/badge/python-3.10%E2%80%933.14-blue" alt="Python 3.10–3.14">
+  <a href="LICENSE"><img src="https://img.shields.io/github/license/duceppemo/seqcounter" alt="License"></a>
+  <a href="https://github.com/duceppemo/seqcounter/wiki"><img src="https://img.shields.io/badge/docs-wiki-informational" alt="Documentation"></a>
+</p>
+
+Count how many sequencing reads (fastq) or contigs (fasta) contain each sequence of a reference file, with
+[Seal](https://sourceforge.net/projects/bbmap/) from BBTools. Point it at a reference fasta file and a sample, or a
+folder of samples, and it writes one table of counts per reference sequence and per sample.
+
+Any set of short sequences works: spoligotype spacers, marker or resistance genes, primer or probe sequences,
+plasmid or phage fragments, barcodes, spike-ins. seqcounter does not interpret them; it counts them, reproducibly
+and in the same way for reads and for assemblies.
+
+```
+$ seqcounter -ref refs.fasta -i samples/ -o results/
+Reference   assembly  no_match  paired  single_end
+absent      0         0         0       0
+present     1         0         86      34
+twice       1         0         154     65
+```
+
+## Features
+* **Reads or assemblies**: single-end or paired-end fastq, or fasta, gzipped or not. Both mates of a pair are
+  counted, once each.
+* **Batch mode**: point it at a folder; fastq and fasta files are found recursively and R1/R2 files paired
+  automatically. Several samples can be counted at the same time.
+* **Both strands, and mismatches**: a reference is found on either strand, and with up to `-m` mismatches per k-mer
+  (one by default).
+* **Three reports**: the counts table (`refseqcounts.tsv`, the table the first version of seqcounter wrote, with the
+  command that produced it on its first line), a per-sample summary of how much was read and how much matched, and a
+  JSON report with the parameters, the software versions and the counts.
+* **Nothing is hidden**: references Seal cannot find (shorter than the k-mer size), names that would collide, and
+  samples that failed are reported, not silently dropped. A failed sample leaves the others readable.
+* **Nothing to configure**: no database, no index, no reference genome. Only the Python standard library, plus
+  BBTools.
+* **Tested**: 105 tests, 100% coverage, end to end on synthetic references, genomes and reads with known counts.
+
+## Installation
+```
+conda create -n seqcounter -c conda-forge -c bioconda bbmap python">=3.10"
+conda activate seqcounter
+pip install seqcounter
+```
+BBTools (which provides `seal.sh`) is not on PyPI, so it comes from bioconda; seqcounter itself needs nothing else.
+See [Installation](https://github.com/duceppemo/seqcounter/wiki/Installation) for other options, including running
+it from a clone with `python seqcounter.py`, as the first version did.
+
+## Quick start
+```
+seqcounter -ref refs.fasta -r1 sample_R1.fastq.gz -r2 sample_R2.fastq.gz -o results/  # Paired-end reads
+seqcounter -ref refs.fasta -r1 sample.fastq.gz -o results/                            # Single-end reads
+seqcounter -ref refs.fasta -r1 assembly.fasta -o results/                             # An assembly
+seqcounter -ref refs.fasta -i samples/ -o results/ -p 4                               # A folder, 4 at a time
+```
+`-ref` is a fasta file of the sequences to count: each one gives a row, named after the first word of its `>` line.
+The counts are printed and saved in `results/`. To see it work on data with known counts, run
+`bash examples/run_example.sh`, which generates its own references and samples.
+
+**What a count is**: the number of reads, or contigs, that contain the reference sequence, not the number of copies
+it has in them. A reference present twice in a single-contig assembly is counted once; in reads, about twice as many
+reads match it. See [How it works](https://github.com/duceppemo/seqcounter/wiki/How-it-works).
+
+## Documentation
+The **[wiki](https://github.com/duceppemo/seqcounter/wiki)** covers
+[installation](https://github.com/duceppemo/seqcounter/wiki/Installation),
+[usage and options](https://github.com/duceppemo/seqcounter/wiki/Usage),
+[output files](https://github.com/duceppemo/seqcounter/wiki/Output-files),
+[how it works](https://github.com/duceppemo/seqcounter/wiki/How-it-works),
+[troubleshooting](https://github.com/duceppemo/seqcounter/wiki/Troubleshooting) and the
+[FAQ](https://github.com/duceppemo/seqcounter/wiki/FAQ).
+
+## Citing
+If you use seqcounter, please cite it and BBTools, which does the counting:
+
+> Duceppe M-O. seqcounter: count the occurrences of reference sequences in sequencing reads or genome assemblies.
+> https://github.com/duceppemo/seqcounter
+
+> Bushnell B. BBTools. https://sourceforge.net/projects/bbmap/
+
+GitHub's **"Cite this repository"** button gives the citation in APA and BibTeX formats, from
+[`CITATION.cff`](CITATION.cff).
+
+## Related
+* [spoligotyper](https://github.com/duceppemo/spoligotyper) — *in silico* spoligotyping of the *M. tuberculosis*
+  complex. It counts the 43 spacers the way seqcounter counts any reference, and then reads them as a spoligotype,
+  with the species, the lineage and a PDF report.
+* [brucetyper](https://github.com/duceppemo/brucetyper) — *in silico* Bruce-ladder, Suis-ladder, AMOS and RB51
+  junction PCR for *Brucella*, from assemblies.
+
+## Contributing
+Bug reports, questions and pull requests are welcome: see [CONTRIBUTING.md](CONTRIBUTING.md).
+
+## Author
+Marc-Olivier Duceppe, Canadian Food Inspection Agency (CFIA): marc-olivier.duceppe@inspection.gc.ca
+
+## License
+[MIT](LICENSE)

seqcounter-1.0.0/pyproject.toml

@@ -0,0 +1,72 @@
+[build-system]
+requires = ["setuptools>=77"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "seqcounter"
+dynamic = ["version"]
+description = "Count the occurrences of reference sequences in sequencing reads or genome assemblies"
+readme = "README.md"
+license = "MIT"
+license-files = ["LICENSE"]
+authors = [{name = "Marc-Olivier Duceppe", email = "marc-olivier.duceppe@inspection.gc.ca"}]
+requires-python = ">=3.10"
+keywords = ["bioinformatics", "kmer", "counting", "reference sequences", "fastq", "fasta", "wgs", "BBTools", "Seal"]
+classifiers = [
+    "Development Status :: 5 - Production/Stable",
+    "Environment :: Console",
+    "Intended Audience :: Science/Research",
+    "Operating System :: POSIX :: Linux",
+    "Operating System :: MacOS",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+    "Programming Language :: Python :: 3.14",
+    "Topic :: Scientific/Engineering :: Bio-Informatics",
+]
+# Seal, from BBTools (https://sourceforge.net/projects/bbmap/), must also be installed: it is not available from
+# PyPI. seqcounter itself needs nothing but the Python standard library.
+dependencies = []
+
+[project.optional-dependencies]
+test = ["pytest>=7", "pytest-cov>=5"]
+
+[project.urls]
+Homepage = "https://github.com/duceppemo/seqcounter"
+Documentation = "https://github.com/duceppemo/seqcounter/wiki"
+Changelog = "https://github.com/duceppemo/seqcounter/wiki/Changelog"
+Issues = "https://github.com/duceppemo/seqcounter/issues"
+
+[project.scripts]
+seqcounter = "seqcounter.cli:main"
+
+[tool.setuptools]
+packages = ["seqcounter"]
+
+[tool.setuptools.dynamic]
+# "seqcounter.__init__" rather than "seqcounter": the root seqcounter.py shim has the same name as the package,
+# and setuptools resolves the plain name to that file, which holds no version.
+version = {attr = "seqcounter.__init__.__version__"}
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+
+[tool.coverage.run]
+source = ["seqcounter"]
+branch = true
+
+[tool.coverage.report]
+show_missing = true
+
+[tool.ruff]
+line-length = 120
+target-version = "py310"
+
+[tool.ruff.lint]
+select = ["E", "F", "W", "I", "UP", "B", "SIM"]
+ignore = ["UP031", "UP032"]  # str.format() is used throughout
+
+[tool.ruff.lint.per-file-ignores]
+"tests/*" = ["E501"]

seqcounter-1.0.0/seqcounter/__init__.py

@@ -0,0 +1,3 @@
+"""Count occurrences of reference sequences in sequencing reads or genome assemblies."""
+
+__version__ = '1.0.0'

seqcounter-1.0.0/seqcounter/__main__.py

@@ -0,0 +1,6 @@
+"""Allow running the command with "python -m seqcounter"."""
+
+from .cli import main
+
+if __name__ == '__main__':
+    main()

seqcounter-1.0.0/seqcounter/cli.py

@@ -0,0 +1,184 @@
+"""Command line entry point."""
+
+import logging
+import os
+import re
+import shlex
+import sys
+import time
+from argparse import ArgumentParser, ArgumentTypeError, RawDescriptionHelpFormatter
+from datetime import datetime
+from pathlib import Path
+
+from . import __version__
+from .counts import count_samples, print_table, run_info, write_reports
+from .errors import SeqCounterError
+from .references import load_references
+from .samples import find_samples, single_sample
+from .seal import DEFAULT_KMER_SIZE, DEFAULT_MISMATCHES, auto_memory, check_seal
+
+log = logging.getLogger('seqcounter')
+
+# Expected errors are reported without a traceback
+USER_ERRORS = (SeqCounterError, OSError)
+
+EPILOG = '''examples:
+  seqcounter -ref refs.fasta -r1 sample_R1.fastq.gz -r2 sample_R2.fastq.gz -o results/
+  seqcounter -ref refs.fasta -r1 sample.fastq.gz -o results/
+  seqcounter -ref refs.fasta -r1 assembly.fasta -o results/
+  seqcounter -ref refs.fasta -i folder_of_fastq_and_fasta/ -o results/
+
+documentation: https://github.com/duceppemo/seqcounter/wiki'''
+
+
+def setup_logging(verbose=False):
+    """Configure the package logger only, so importing seqcounter never alters the root logger."""
+    for handler in list(log.handlers):
+        log.removeHandler(handler)
+    handler = logging.StreamHandler(sys.stderr)
+    handler.setFormatter(logging.Formatter('%(asctime)s %(levelname)-7s %(message)s', datefmt='%H:%M:%S'))
+    log.addHandler(handler)
+    log.setLevel(logging.DEBUG if verbose else logging.INFO)
+
+
+def positive_int(value):
+    try:
+        value = int(value)
+    except ValueError:
+        raise ArgumentTypeError('"{}" is not an integer'.format(value)) from None
+    if value < 1:
+        raise ArgumentTypeError('must be >= 1')
+    return value
+
+
+def java_memory(value):
+    if not value[:-1].isdigit() or value[-1].lower() not in 'mg':
+        raise ArgumentTypeError('"{}" is not a Java memory size such as 500m or 2g'.format(value))
+    return value
+
+
+def available_cpus():
+    try:
+        return len(os.sched_getaffinity(0))  # Respects cgroup/taskset limits on Linux (e.g. SLURM jobs)
+    except AttributeError:
+        return os.cpu_count() or 1
+
+
+def format_elapsed(seconds):
+    """A readable duration, as the first version of seqcounter printed it: "1h2m3s"."""
+    minutes, seconds = divmod(int(round(seconds)), 60)
+    hours, minutes = divmod(minutes, 60)
+    days, hours = divmod(hours, 24)
+    periods = [('d', days), ('h', hours), ('m', minutes), ('s', seconds)]
+    return ''.join('{}{}'.format(value, name) for name, value in periods if value) or '0s'
+
+
+def build_parser():
+    max_cpu = available_cpus()
+    parser = ArgumentParser(prog='seqcounter', formatter_class=RawDescriptionHelpFormatter, epilog=EPILOG,
+                            description='Count the occurrences of reference sequences in sequencing reads (fastq)\n'
+                                        'or genome assemblies (fasta), with Seal from BBTools.')
+    inputs = parser.add_argument_group('input (one sample with -r1, or a folder of samples with -i)')
+    source = inputs.add_mutually_exclusive_group(required=True)
+    source.add_argument('-r1', '--r1', metavar='FILE',
+                        help='Reads (single-end or R1 of paired-end) or assembly. fastq or fasta, gzipped or not.')
+    source.add_argument('-i', '--input', metavar='FOLDER',
+                        help='Folder of fastq and fasta files, searched recursively. Each fasta file, single-end '
+                             'fastq file or pair of R1/R2 fastq files is a sample.')
+    inputs.add_argument('-r2', '--r2', metavar='FILE',
+                        help='R2 reads, for paired-end fastq files.')
+    inputs.add_argument('-ref', '--ref', metavar='FILE', required=True,
+                        help='Fasta file of the reference sequences to count, gzipped or not. One row per sequence '
+                             'is reported, named after the first word of its ">" line.')
+    inputs.add_argument('-s', '--sample', metavar='NAME',
+                        help='With -r1: sample name used in the reports. Default: the input file name, without '
+                             'extension and read suffix (_R1, _1, ...).')
+
+    output = parser.add_argument_group('output')
+    output.add_argument('-o', '--output', metavar='FOLDER', required=True,
+                        help='Folder to hold the reports. Created if needed.')
+
+    counting = parser.add_argument_group('counting')
+    counting.add_argument('-k', '--kmer-size', metavar='N', type=positive_int, default=DEFAULT_KMER_SIZE,
+                          help='K-mer size used to find the reference sequences. A reference shorter than k cannot '
+                               'be found. Default: {}.'.format(DEFAULT_KMER_SIZE))
+    counting.add_argument('-m', '--mismatch', metavar='N', type=int, default=DEFAULT_MISMATCHES,
+                          help='Mismatches allowed between a reference k-mer and the input (maximum Hamming '
+                               'distance). Default: {}.'.format(DEFAULT_MISMATCHES))
+
+    other = parser.add_argument_group('performance and other options')
+    other.add_argument('-t', '--threads', metavar='N', type=positive_int, default=max_cpu,
+                       help='Number of threads. Default: all available ({}).'.format(max_cpu))
+    other.add_argument('-p', '-j', '--parallel', '--jobs', metavar='N', type=positive_int, default=1, dest='parallel',
+                       help='With -i: number of samples counted at the same time, sharing the threads. Default: 1.')
+    other.add_argument('--memory', metavar='SIZE', type=java_memory,
+                       help='Memory for Seal (Java heap size), per job, e.g. 2g. Default: from the size of the '
+                            'reference sequences, within 85%% of the machine.')
+    other.add_argument('--verbose', action='store_true', help='Show debug messages, including the Seal command.')
+    other.add_argument('-v', '--version', action='version', version='%(prog)s ' + __version__)
+    return parser
+
+
+def check_arguments(parser, args):
+    if args.input and (args.r2 or args.sample):
+        parser.error('-r2 and --sample are only used with -r1')
+    if args.r1 and args.parallel > 1:
+        parser.error('--parallel is only used with -i')
+    if args.mismatch < 0:
+        parser.error('--mismatch cannot be negative')
+    if args.mismatch >= args.kmer_size:
+        parser.error('--mismatch ({}) must be smaller than --kmer-size ({})'.format(args.mismatch, args.kmer_size))
+    if args.sample is not None and (not args.sample.strip() or '/' in args.sample or args.sample in ('.', '..')):
+        parser.error('--sample must be a name, not a path: "{}"'.format(args.sample))
+    if args.sample is not None and re.search(r'[\t\r\n]', args.sample):
+        parser.error('--sample cannot contain tabs or line breaks (they would break the table)')
+
+
+def main(argv=None):
+    parser = build_parser()
+    args = parser.parse_args(argv)
+    check_arguments(parser, args)
+    setup_logging(args.verbose)
+    command = shlex.join([parser.prog, *(sys.argv[1:] if argv is None else argv)])
+    log.debug('seqcounter %s: %s', __version__, command)
+    started = datetime.now().astimezone()
+    start = time.monotonic()
+    output = Path(args.output).expanduser()
+    try:
+        check_seal()
+        references = load_references(args.ref, args.kmer_size)
+        if args.input:
+            samples = find_samples(args.input, exclude=[output])
+            log.info('%d sample(s) found in %s', len(samples), args.input)
+        else:
+            samples = [single_sample(args.r1, args.r2, name=args.sample)]
+        if any(s.file_type == 'fasta' for s in samples):
+            log.info('fasta input: the count of a reference sequence is the number of sequences (contigs) that '
+                     'contain it, not the number of copies it has in them.')
+        jobs = min(args.parallel, len(samples))
+        memory = args.memory or auto_memory(sum(r.length for r in references), jobs=jobs)
+        results = count_samples(samples, args.ref, jobs=jobs, threads=args.threads, kmer_size=args.kmer_size,
+                                mismatch=args.mismatch, memory=memory)
+        run = run_info(command, {'Reference': os.path.abspath(args.ref),
+                                 'Input': ' '.join(os.path.abspath(f) for f in (args.input, args.r1, args.r2) if f),
+                                 'Output folder': str(output.resolve()),
+                                 'K-mer size': args.kmer_size, 'Mismatches': args.mismatch,
+                                 'Threads': args.threads, 'Parallel samples': jobs, 'Seal memory': memory}, started)
+        rows, reports = write_reports(results, references, output, command=command, run=run)
+    except USER_ERRORS as e:
+        log.error('%s', e)
+        log.debug('Traceback:', exc_info=True)  # Shown with --verbose
+        sys.exit(1)
+    except KeyboardInterrupt:
+        log.error('Interrupted')
+        sys.exit(130)
+
+    # The counts go to stdout, the messages to stderr: "seqcounter ... > counts.tsv" works
+    print_table(rows)
+    for report in reports:
+        log.info('Report saved in %s', report)
+    log.info('Done in %s', format_elapsed(time.monotonic() - start))
+    failed = [r.sample for r in results if r.error]
+    if failed:
+        log.error('%d of %d sample(s) failed: %s', len(failed), len(results), ', '.join(failed))
+        sys.exit(1)